Hydrogen atom transfer reactions are fundamental to a large class of enzymes that catalyze a diverse array of important metabolic reactions in which carbon-based radicals are key intermediates. Radical abstraction of a non-acidic hydrogen atom from the substrate is the key activation step in a variety of unusual and chemically difficult transformations. In many enzymes 5'-deoxyadenosyl radical, generated from either adenosylcobalamin or reduction of S-adenosylmethionine, serves as the cofactor for hydrogen abstraction; in others a protein-based radical, often a cysteinyl radical, serves as the radical cofactor. Compared to their reactivity in free solution, free radicals generated in enzymes appear to be stabilized to a remarkable degree. The mechanisms by which enzymes generate and stabilize reactive free radicals remain poorly understood. Our aim is to investigate in detail how hydrogen atom transfer, the key step in substrate activation, is catalyzed with the goal of illuminating the general principles by which enzymes catalyze radical reactions. We will focus on understanding hydrogen atom transfer in two radical enzymes that serve as model systems. Glutamate mutase is an adenosylcobalamin-dependent enzyme in which hydrogen atom transfer serves to generate a substrate radical that then undergoes a carbon skeleton rearrangement. Benzylsuccinate synthase is a glycyl-radical enzyme in which hydrogen transfer from toluene to an active site cysteine is a key mechanistic feature. We will integrate experimental measurements on enzymes and non-enzymatic model reactions with computational studies of these reactions to provide a framework within which to understand the differences between enzyme-catalyzed hydrogen atom transfer reactions and non-enzymatic reactions. Among the experiments we will conduct are kinetic isotope effect measurements that aim to determine to what extent quantum tunneling of the migrating hydrogen atom is important in the reaction catalyzed by glutamate mutase. We will compare these results with those obtained for mutant enzymes and model, non- enzymatic B12 reactions. By modeling both enzymatic and non-enzymatic reactions using state-of-the-art QM/MM computational methods we will gain insights into catalysis by the enzyme, and, in particular, whether the enzyme actually increases the amount of quantum tunneling in a reaction to enhance catalysis. We will test whether enzyme-catalyzed hydrogen atom transfer reactions can rationalized by linear free energy relationships such as the extended Evans-Polanyi equation and Hammett analysis that successfully predict the rates of a large number of non-enzymatic hydrogen transfer reactions based on simple empirically determined parameters. We will measure the rates of hydrogen transfer in benzylsuccinate synthase and glutamate mutase when reacted with substrate analogs containing substituents capable of stabilizing the resultant substrate radicals by different amounts. These experiments should provide insights into the relative importance of enthalpic, steric and polar effects in enzyme-catalyzed hydrogen transfer reactions.